This is a revised application of Glucocorticoid Antagonist Treatment of Alcohol Use Disorder (AUD). We addressed approach and feasibility concerns by providing results from our recently completed proof-of-concept (POC) human laboratory study of mifepristone (600 mg/d) in 50 non treatment-seeking men and women with AUD, and by incorporating the admission criteria and safety parameters included in our IND (#114497) for the mifepristone POC study, which are directly relevant to the aims of the present project. The development of more effective and better tolerated pharmacotherapies for AUD that have greater acceptability to patients and clinicians is a NIAAA priority. The hypothesis under test in this proposal is that antagonism of the glucocorticoid receptor is an effective treatment fr AUD. The glucocorticoid receptor is integral to the brain stress response associated with onset, maintenance and relapse in AUD. We propose to test the efficacy of a potent glucocorticoid receptor antagonist, mifepristone that acts to rapidly restore and maintain normal tone in the brain stress system. Our proposed clinical trial of mifepristone is based on positive results obtained with mifepristone in pre-clinical and human laboratory models of protracted abstinence. In our POC study, 1-week of treatment with mifepristone 600 mg/d significantly reduced cue- and stress-induced craving in the lab; significant decreases in drinking, craving and GGT, and improvement on tests of learning, memory and executive function persisted for 1-month post treatment relative to placebo in 50 non treatment seekers with AUD. We propose to extend our POC study results to a treatment seeking sample, with 1-week of mifepristone to increase rate of initial response by re-setting the HPA axis and 8 weeks of counseling to consolidate and sustain mifepristone effects in 150 men and women with DSM-V AUD of e moderate severity. Based on the significant association between better drinking outcome and higher mifepristone plasma concentration found in our 600 mg POC study in conjunction with no safety or tolerability concerns, we propose to conduct a 3-arm dose-ranging study with randomization to double-blind treatment with 600 or 1200 mg/d of mifepristone or matched placebo in order to identify optimal dosing for AUD. Our Primary Aim is to determine if mifepristone treatment is associated with significant improvement in drinking outcomes in AUD relative to placebo. We hypothesize mifepristone will be associated with: 1.) significant linear dose-related reductions in the number of heavy drinking days per week and the number of drinks consumed per week, and 2.) a significant linear dose- related increase in rates of no heavy drinking over the 8-week study relative to placebo. Our Secondary Aims are to determine if mifepristone is associated with a significantly greater reduction in craving and improvement in neurocognitive functioning than placebo. Our Exploratory Aim is to determine whether mifepristone effects on drinking outcome are predicted by polymorphisms in the FKBP5 gene, an important regulator of the glucocorticoid complex and cortical response, to provide more targeted and effective treatment of AUD.